Determining Optimum Acceptance Sample Size—A Second Look

摘要

What is the optimum acceptance sample size (n)? Recent research by Gharaibeh et al. found that,when incoming lots are anticipated to be high quality, the optimum n is relatively small (n = 3for many acceptance quality characteristics). The finding was received with some skepticism,and many questions were raised regarding the optimization model and underlying assumptions.This paper summarizes follow-up research that more closely examined the optimization modelelements and assumptions in an attempt to confirm or refute Gharaibeh's finding. With asomewhat modified model and assumptions based where possible on actual data, the conclusionsreached by Gharaibeh stay basically the same and several new conclusions are drawn. Somecautions are presented regarding the state highway agency use of small sample sizes. Agenciesare encouraged to perform their own optimization calculations. By doing so, they will gain abetter understanding of their acceptance plan systems and associated costs, and have greaterconfidence in applying economic decision analysis principles to minimize expected costs andoptimize statistical acceptance risks.